[Bug target/43743] Dumb use of SSE regs for MMX operation

ubizjak at gmail dot com gcc-bugzilla@gcc.gnu.org
Mon Apr 19 12:54:00 GMT 2010



------- Comment #2 from ubizjak at gmail dot com  2010-04-19 12:53 -------
(In reply to comment #0)

> If I take off -msse2, I get this:
> 
>         movd    (%eax), %mm0
>         movd    4(%eax), %mm1
>         punpcklbw       %mm1, %mm0

The code above is produced _without_ SSE (-mno-sse), since with -mno-sse2, you
will get:

        movss   (%eax), %xmm0
        movlps  %xmm0, -8(%ebp)
        movss   4(%eax), %xmm0
        movq    -8(%ebp), %mm0
        movlps  %xmm0, -16(%ebp)
        punpcklbw       -16(%ebp), %mm0

So, effectively the same code as with SSE2.


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2010-04-19 12:53:43
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43743



More information about the Gcc-bugs mailing list